Hi 👋🏻
Entities are the backbone of search—old and new. Traditional engines annotate text with entities to rank, disambiguate, and power knowledge panels. LLMs compress the world into vectors, then still rely on entities to ground answers, retrieve or generate approximate potential queries, call tools, and avoid nonsense. If you aren’t modeling your world as entities, you’re shipping ambiguity.
- LLMs synthesize prose, but the “facts” underneath are still entity-level claims (thing → attribute → value). Your content wins when those claims are precise, verifiable, and connected.
- Classic information retrieval + knowledge graphs now sit beside vector stores and RAG. The join key is the entity.
- Context is continuous. Sessions, devices, and assistants carry state about which entity a user means—and which attribute they care about right now.
Think in entities, not pages. Treat your site like a product graph, not a blog. Give every important entity a canonical home, structured attributes, relationships, and proofs.
Then make formats that line up with how engines and LLMs retrieve: cards, specs, comparisons, lists. Go multimodal - video, image, podcasts, forums - wherever your audience is, meet them there.
Ensure you continuously surface new, unique information and angles to your audience. LLMs reward content freshness, and have the potential to seriously amplify unique perspectives. Nothing new in terms of tactics, but vital stuff.
Entities and entity linking are used throughout the AI search process. Here are just some of the ways this is done:
- LLMs detect entities in the user query.
- LLMs recognise entities in signals (prior queries, location, device, behaviour).
- LLMs use entity-based search in docs/passages to form a custom corpus, from which an answer is generated
- Some AI search systems cite/verify via entity-grounded sources
- Some AI search systems (like AI mode) can expand a user query using entity-based linking
Keywords describe text. Entities describe reality. Traditional search and LLMs both reward the latter.
Featured course 🌟
SEMANTIC AI-powered keyword research
We're hard at work on the updates to this course - two new modules are in development, in which we'll focus on Synthetic Queries and Brand Entities for LLM-based search optimisation.
|
|
|
Related to entities in the course, we currently cover things like:
- Importance of Entities and Knowledge Graphs for traditional search
- EAV (Entity-Attribute-Variable) model in practice in the keyword research process
- How to find semantically linked entities and extract entity information with the NLP API and Knowledge Graph APIs by Google Cloud
But also a TON more 🔥 10+ hours of content; 20+ tools, checklists, scripts, and other resources. 🔥
✨ Use code COMMUNITY30 for 30% off at checkout - that's our little thank you for being part of the MLforSEO community
Heads-up 💸 the price of this course will increase when the new modules drop—lock in now.
We also had some positive feedback on this very course - see what Sr. SEO and Content Strategist at Piktochart, Vinnie Wong, had to say about the course:
65+ forward-thinking marketers are already taking our courses 💜
Launched our experts hub
We recently revamped our experts hub, bringing you a list of 41 vetted experts in the AI/ML SEO space.
We've also added a few of our favourite pieces of content by each one, so you can kickstart your learning.
If you're not seeing your favourite content creator in the directory, let us know who else should be on this list.
We're working on the MLforSEO tools library next, so if you have a tool you'd like to feature, submit it via this form.
recent discussions from our slack community 💬
Here's some of the recent discussions there:
Or check out a small sample of the resources shared:
Join over 600 AI/ML-interested marketers on our Slack community (if you haven't already) to stay up to date with discussions on AI/ML automation in SEO and marketing.
Happy learning! ✨
Lazarina
